Lessons learned from mine (2’x3’ working area), at least 1" tubing. Mid span supports would help. The bigger you go the more you’ll see mid span rigidity issues, hence most bigger builds are the lowrider variety. Dust collection is a must.
I had thought about solid rods for rails, but someone had mentioned the steppers aren’t going to handle moving that much mass. Could possible use solid for the outer rails as they don’t move, but at longer lengths, will still get flex in mid span (with tubing). A lowrider might be the better way to go with the size you want to build. JMHO
I have a 2’ by 3’ Primo, and it’s too big. It is usable at that size, but I get some issues near the middle of the long span if I’m not super careful with the CAM. The LowRider is much better about this.
I’ll be adjusting the Primo to probably 19" by 25" (Currently 37" by 25") and I’m considering 19" square, in order to be able to handle work with aluminum.
Beyond using sturdy steel pipe/rods for the perimeter rails. What else are people doing to make their MPCNCs even more rigid?
Built a LowRider instead of MPCNC for some of the reasons being mentioned here. But there’s lots of MPCNCs out there though, am curious because MPCNC still seems like a good gantry setup for smaller CNC and other projects…
Is there enough gap/clearance on underside of the rails (between the truck bearing wheels) to weld a plate along the EMT, to help minimize rails flexing up/down?
If welding isn’t an option, is there enough clearance on underside of outer most exterior EMT rails to drill keyholes every 200mm or so (3D print a hold down alignment jig, then use stock MPCNC to very slowly drill keyholes). Then fasten bolts or threaded rod that connect to the table via 3D printed mount with captive bolt, that rotates to adjust height/tension, screw/bolt mount to the table. Something like MPCNC’s Z axis nut traps, but for perimeter rails. At some point, might be better use of time to buy a linear rail guide, design remix/mods would be needed though.
Noticed @Waspjr just created couple of topics with some ideas:
IMO, it is not the rails that are the problem, it is the gantry tubes. These can’t be supported, because they have to move, pretty much by definition. Maybe the rails if I were using conduit, but the 1" steel that I’m using seems pretty stable for the rails. It’s the long gantry rail that starts to chatter under stress. The lateral stresses on the rails are split between the 2 rails, but the gantry tubes need to take all of the stresses by themselves.
When I get chatter, I can feel the gantry tubes vibrating, but the rails are generally not bad.
Mid-span supports help with sag, but probably don’t really help much with rigidity against lateral stress. Countering sag is not insignificant, but not the main problem that I experience.
Could you detect and reduce lateral stresses that eventually result in chatter, if the belt had a tension gauge sensor that led to the controller dynamically adjusting feedrate? Maybe detecting chatter sounds/vibrations would be simpler. This doesn’t help with increasing speeds. Maybe the MPCNC Gantry needs more EMT and Struts
Maybe… but the worst cases for me are no warning, like plunging into the material for a small hole that ends up being a bigger hole than I called for. That seems to be where the worst chatter happens. I found that I can minimise it if I know it’s likely to happen by holding the long rail in my hands, but that doesn’t always stop it.
I recently made some cabinet doors. Some mid span flex/chatter left unexpected tool marks in the middle, which needed to be filled and sanded before painting. If i made more shallower passes, there would’ve been virtually no issues.
At roughly 50 minutes a pass for 4-6 passes, it takes too much time to justify making each cabinet door. While it is doable, the amount of time to do so becomes a factor. If that makes any sense…